138 



OCEAN ATMOSPHERIC -ELECTRIC RESULTS 



shows only a slight tendency to conform. One must 

 recognize, however, that the five days making up the 

 Septemlwr group are widely scattered over a period 

 when very large changes in conductivity occurred. The 

 diurnal variation of negative conductivity may, perhaps, 

 be better viewed on Greenwich time, since the character 

 of variation is such as to suggest a relation with the 

 variation in potential-gradient. For Investigation of this 

 point, data were assembled for the preparation of the 

 graphs shown in figure 5. 



For figure 5, the registrations of conductivity were 

 examined in relation to registrations of potential gradi- 

 ent, the selection of days being based on the availability 

 of undisturbed days on which both elements were satis- 

 factorily recorded, and, equally important, on the avail- 

 ability of groups of successive days of data for which 

 alternate registration of positive and negative conductiv- 

 ity indicated a homogeneity of conditions. Data for the 

 period October 5 to November 12 were selected and in- 

 cluded twenty-one days of potential-gradient, thirteen 

 days of positive conductivity and twelve days of negative. 

 Fewer days of potential-gradient than of conductivity data 

 were available because no registrations of this element 

 were obtained on October 12, 17, 18, 22, and November 

 4, whereas on one day, November 11, potential-gradient 

 but no conductivity data were available. The homogene- 

 ity of the material, however, made it appear desirable 

 to include these days and thus increase the amount of 

 data. No September data were used because much loss 

 of record occurred in that month and the available com- 

 plete days were scattered and covered an unusually large 

 range of values. 



The graphs in figure 5, reading from top to bottom, 

 represent (a) positive conductivity, (b) negative conduc- 

 tivity, (c) total conductivity, (d) the ratio of positive to 

 negative conductivity, and (e) the potential-gradient. The 

 diurnal variations in positive and negative conductivity 

 show no important departures from the graphs already 

 given in figures 3 and 4. The range from maximum to 

 minimum in positive conductivity is 9 per cent and in 

 negative conductivity 12 per cent. The difference in 

 time between the maximum in positive conductivity and 

 the minimum in negative causes the total conductivity to 

 be low during the period from 18h to 22h GMT, but 

 otherwise the total conductivity is essentially constant. 

 The ratio of the two conductivities, on the other hand, 

 undergoes a considerable variation and the character of 

 that variation is very similar to the variation in potential- 

 gradient, having its maximum at 18h to 19h GMT, with 

 gradual rise to the maximum and abrupt falling off there- 

 after. Similar agreement will be found in cruise VI data 

 which will be examined later. The negative conductivity 

 varies in a manner opposite to that of the potential -gra- 

 dient, suggesting that even a small diurnal change in 

 field in fair weather may produce an appreciable "elec- 

 trode effect," causing the negative small-ion content of 

 the atmosphere near the earth's surface (and therefore 

 the negative conductivity) to diminish as the potential- 

 gradient increases. Since the variation in potential-gra- 

 dient proceeds according to universal time (Greenwich 

 time), as discussed in the preceding paper, rather than 

 local time, this would imply that the variation in nega- 

 tive conductivity should do likewise. 



Whether the diurnal variation in either or both con- 

 ductivities proceeds according to local or to universal 

 time cannot be decided from the three months of record- 

 ed data, because of the restricted distribution in longitude. 



Therefore, comparisons with data obtained by manual 

 observations earlier on cruise VH and data obtained in 

 1921 on cruise VI are of some interest. Diurnal varia- 

 tion data were obtained by manual observations on 

 cruise VH for complete twenty-four hour periods on 

 twenty -three days between July 29, 1928 and July 28, 

 1929. Measurements of positive conductivity were made 

 on seventeen of these days and of negative on six days, 

 as shown in the table on pages 103 to 112. Of the seven- 

 teen days of positive conductivity, three were obtained 

 in the north central Atlantic Ocean at approximately 320° 

 east longitude, eight in the southeastern Pacific Ocean 

 between 237° and 280° east, one in the north central Pa- 

 cific near 220° east, and five in the northwestern Pacific 

 between 144° and 185° east. Only the eight-day group 

 (reduced to seven days by discarding one disturbed day) 

 offers enough material for accomparison with the gr^h 

 for positive conductivity in figure 3. The comparison is 

 made in figure 6 where the agreement perhaps is some- 

 what better on a Greenwich time basis than on a local 

 time basis. Thus, for cruise VII, two groups of data 

 separated about 50 degrees in longitude, the measure- 

 ments in one obtained manually and in the other by reg- 

 istration, give comparable results for the character of 

 the diurnal variation in positive conductivity. The graphs 

 in figure 6 are based on a very considerable part of the 

 total days of data for positive conductivity obtained on 

 cruise VII; forty- seven days of data were obtained of 

 which seven were disturbed by bad weather. Of the re- 

 maining forty days, thirty --75 per cent — were used in 

 figure 6. There is, therefore, a great preponderance of 

 cruise Vn data favoring a diurnal variation in positive 

 conductivity of the character shown in figure 6. 



Turning to examination of positive conductivity 

 measurements made on cruise VI, the diurnal variation 

 is found to be quite different. This feature is not, how- 

 ever, the only one in the earlier work which merits 

 some discussion. Between April 9 and August 31, 1921, 

 on cruise VI, manual measurements were made of both 

 positive and negative conductivity (and simultaneously of 

 potential-gradient) for complete twenty-four-hour periods 

 on ten days in the central Pacific Ocean in the same lon- 

 gitude range, 188° to 235° east, as that in which the data 

 in figure 5 were obtained. On each of the ten days the 

 measurement of positive conductivity was alternated 

 from hour to hour with measurement of negative conduc- 

 tivity, thus giving twelve measurements through each day 

 of each sign of conductivity. One day, July 29 to 30, was 

 disturbed by bad weather, leaving nine days of data from 

 which to prepare the graphs in figure 7. In this figure, 

 the sequence of graphs from top to bottom is the same 

 as in figure 5. 



The character of the diurnal variation of positive 

 conductivity in figure 7 shows no similarity to that in 

 figure 5. Instead, it is very much like the variation in 

 negative conductivity for cruise VI, and the total conduc- 

 tivity for the cruise consequently shows a similar varia- 

 tion. The completely different character of diurnal vari- 

 ation in positive conductivity for cruises VI and vn would 

 seem to indicate that purely local factors were operating 

 on one or both of these two occasions eight years apart, 

 to give the results obtained. Careful examination of the 

 results from both cruises has revealed no reason for 

 questioning the validity of either group of material, and 

 the question as to what brought about the quite different 

 diurnal variation characteristics must remain unanswered 

 for the present. The negative conductivity exhibits some 



